Why did they wait so long to finally abandon blower style cooling

>>59886722
dunno but the 550 is looking to be a good drop-in GPU if it runs off the PCI lane or has one 6-pin

Because AMD has been listening to everything people bitch about.

>AMD has bad drivers
Fixed
>AMD has bad IPC
Fixed
>AMD has bad power consumption
Fixed
>AMD isn't as good for gaming
Fixed

They may have been behind for a while, but they have been playing major catch up recently. If it continues both Nvidia and Intel are going to get destroyed.

>>59887063
Their Linux drivers need to be on par with Nvidia's proprietary

>>59887093
True, but they have gotten significantly better recently. You can expect around 70-80% of Windows performance.

>>59887063
The drivers still suck and their management program crashes when you try to excess watt man because it scans for all installed games before you can switch to watt man.

>>59887063
The 5xx are rebranded and the power draw is like Maxwells. Thats also why AMD gpus suck on laptops

>>59887176
What's wrong with a rebrand? It's not like it's going to make Vega take any longer. It just means more support for the RX480.

And Polaris has better performance to watt than Maxwell, just not as good as Pascal.

>>59887152
I haven't tried to get any control panel working, I have just installed the GPGPUPRO drivers or whatever. I wasn't aware you could access wattman.

>>59887216
>What's wrong with a rebrand?
I stop read

>>59887216
>>59887176
Power draw aside, I think the rebrand is a good thing. People are buying multiple 1060's and 1050's for each polaris sold becaause of nvidia's marketing, nothing is going to change that until AMD learns how to market and the value of a halo product.
The upside to this rebrand is that when people google for 580 vs 1060 they will see recent benches that show current performance which has improved quite a bit since 480 launch.
Another thing that will hopefully be fixed is the crazy lottery at play with the polaris chips out right now. Some can't clock past 1250, some hit 1500 on air. Some can be undervolted AND overclocked, some can't. When I think of polaris, I think of it running at 1500mhz, just like I think of pascal running at 2100mhz. This should be more universally obtainable.

>>59886737
I want a comfy RX550 for my e-gpu

>>59887093
Their drivers are included in the mainline Kernel, nvidia has to be installed either through packages or a script that causes big reversion problems.

>>59886722
It sucks that the 500 series is just a shitty rebrand. If they increased the rops from 32 to 64 and used GDDR5x VRAM instead, then I think it would be justified.

>>59888875
I agree, they really should have offered a suped up one. There wasn't much of anything stopping them from doing it for the original polaris release. The chip is powerful enough, it just needed better memory and to be a bit bigger. Now I see why they're not doing it, it would step on Vega's toes but then they should have waited for Vega's release to drop these, or copied nvidia and released the vega cards, then these rebrands in Summer.

>>59887216
>Performance to watt

>>59887265
>rebrands are a good thing
AMD, everyone

>>59890840
NVIDIA also does that.

>>59886737
desu I'll be mildly upset if the 550 requires external power. My old 6670 didn't.

>>59886722
when the fuck they will release a X90 card again? I'm tired of the best ones being only X80

kk

>>59886722
talked to amd about this with why they don't package a stock cooler on the 1700x/1800x

Apparently 3rd parties and oems threaten amd and intel to not make their shit better.

>>59892115
they aren't, vega is going to be vega branded

>>59890840
do we know the price? that's really the only way to justify the rebrand this time around.

>>59892324
Only way? The 580 stock are clocked at like 200MHz higher than reference 480, that's a pretty big difference.

Have you people forgot that the reference 480 throttled below its boost clock constantly?

Too bad AMD won't give up on trying to have their own Titan with this Fury/Vega naming shit

The rebrands wouldn't look so bad if they actually slid the rebrands down a tier and put the Vega chips at the 580/590 spots

>>59886722
>Why did they wait so long to finally abandon blower style cooling for reference models?
Reference designs love blowers so much because they assume less about the cooling capabilities of the rest of the system. Because they exhaust out the back, they work well even if the case is under-ventilated. They can also be jammed together in multi-card configurations better, for the same reason, and because blowers have inherently higher static-pressure ratings than ordinary axial fans. (that is, they can suck/blow air through restrictive openings better)

>>59886737
550 is pci powered, 560 is the full 1024sp and I believe has a 6 pin.

>550 compared to 250, the old "dual graphics" card
>550 compared to intel igp
>550 about 750ti performance
>750ti performance is new igp target
Dare I say this might mirror the performance of the new apu igp? Is AMD going to push hybrid xfire on the apu line?

>>59892419
you are thinking stock, base clocks on 480's you can buy now are higher then the rebrand for what, 20 maybe 30 above msrp?

>>59892713
I was hoping for the new APU to be at least 460 levels of power, then again, they might have an A12 again with a higher power GPU

>>59893761
I'm thinking stock because stock benchmarks are still out there and are used as a basis, like for fucks sakes TPU doesn't have a single AMD GPU review since fucking June 2016.

>>59893851
Never really thought about that, the reference 480 frequently dropped to 1200MHz from its 1270 boost in gaming, and the 580 come with a much nicer cooler that won't throttle and boosts to 1340.

Hopefully using GloFo's LPC over LPP improves the overclocking ability of the chips, some 480's could push to 1500 while others barely reached 1380 and it wasn't a shit PCB issue.

>>59893761
Here's the fastest stock rx 480 available
GTR OC: 1344mhz
Nitro+ OC: 1342mhz
>Ref Rx580: 1340mhz
much higher